Naming conventions for model components

Each analytic model is a set of database tables and database views. These tables and views are called model components.

To each component, a canonical name is assigned by using a special naming schema.

The name is generated in the following way:

<Model_name>_<Derived_component_type>_<Sequence_ID>

Where:

<Model_name> is the name of the model that is created by using the naming conventions for models as described in Naming conventions for models.
<Derived_component_type> is the usage type of the component, for example, MODEL, PMML, or INPUT. Any instances of one or more consecutive spaces in the usage type are converted to an underscore (_).
<Sequence_ID> is the sequence ID of the component.
The sequence ID is required if an algorithm stores more than one table or view of the same type because these tables or views are distinguished by the sequence ID. If the sequence ID is not required, it is omitted.

The following example shows a model Sample Model that uses two tables with usage type Column Statistics due to its structure. In this case, the sequence ID is required.

The table names are as follows:

Sample Model_COLUMN_STATISTICS_0
Sample Model_COLUMN_STATISTICS_1